1. Field of the Invention
The present invention relates to a technique of, in an image capture apparatus, correcting blur of a captured image caused by a shake of the image capture apparatus, using a method of deforming an image.
2. Description of the Related Art
In recent years, along with development of a technique of correcting a shake of an image capture apparatus, an image stabilization function of correcting not only blurring of a captured image caused by a camera shake when a user is in a still state but also blurring of a captured image which is generated in shooting while the user is walking has become widespread. When the user shoots an image while walking, in addition to blurring of the captured image in the horizontal and vertical directions, the following blur is generated in the captured image. For example, blur includes rotation of the captured image caused by rotation of the image capture apparatus about the optical axis, and distortion of the captured image into a trapezoid, which is generated when the image capture apparatus tilts with respect to an object.
As a method of correcting various kinds of blur of a captured image which are generated in shooting while walking, there is known a method of calculating the image deformation amount of the captured image and deforming the image so as to cancel the image deformation amount (see Japanese Patent Laid-Open No. 2011-29735).
In the aforementioned conventional example, a deformation amount of a captured image caused by a shake acting on the image capture apparatus is decomposed into deformation components including a translation component (horizontal/vertical), a perspective component (horizontal/vertical), an enlargement/reduction component, a rotation component, and a shear component. Filtering processing and the like are then performed for each component to calculate a projective transformation matrix (homography matrix). This method, however, presents the following problems.
That is, in the aforementioned conventional example, since a correction amount is calculated for each of the deformation components, a calculation amount is very large and processing is complicated. Furthermore, if correction is performed for all the deformation components, an image range which can be finally output becomes small, resulting in degradation in image quality.
For example, FIG. 21 shows an example of image stabilization by image deformation when an image capture apparatus rotates about the optical axis. A solid-line portion in FIG. 21 indicates a captured image. To correct rotation about the optical axis, the captured image is rotated about an image center O. The rotation operation results in a dotted-line image shown in FIG. 21. In this case, if an output image has the same range as that of the captured image, the image has no data at the four corners. Therefore, the output image range is a hatched region shown in FIG. 21. Although correction of rotation about the optical axis has been exemplified in FIG. 21, it is necessary to provide, for each of the deformation components, an image region (to be referred to as extra pixels hereinafter) which cannot be output. Therefore, the image range becomes small, thereby making it difficult to maintain the quality of video having undergone image stabilization.